DSR1-Qwen-32B-scg / train_results.json
moogician's picture
Upload train_results.json with huggingface_hub
e83cdef verified
raw
history blame
202 Bytes
{
"epoch": 6.0,
"total_flos": 17281228603392.0,
"train_loss": 0.3665730852840675,
"train_runtime": 2825.3225,
"train_samples_per_second": 0.195,
"train_steps_per_second": 0.025
}